Kir6.1 improves cardiac dysfunction in diabetic cardiomyopathy via the AKT‐FoxO1 signalling pathway

Abstract Previous studies have shown that the expression of inwardly rectifying potassium channel 6.1 (Kir6.1) in heart mitochondria is significantly reduced in type 1 diabetes. However, whether its expression and function are changed and what role it plays in type 2 diabetic cardiomyopathy (DCM) have not been reported. This study investigated the role and mechanism of Kir6.1 in DCM. We found that the cardiac function and the Kir6.1 expression in DCM mice were decreased. We generated mice overexpressing or lacking Kir6.1 gene specifically in the heart. Kir6.1 overexpression improved cardiac dysfunction in DCM. Cardiac‐specific Kir6.1 knockout aggravated cardiac dysfunction. Kir6.1 regulated the phosphorylation of AKT and Foxo1 in DCM. We further found that Kir6.1 overexpression also improved cardiomyocyte dysfunction and up‐regulated the phosphorylation of AKT and FoxO1 in neonatal rat ventricular cardiomyocytes with insulin resistance. Furthermore, FoxO1 activation down‐regulated the expression of Kir6.1 and decreased the mitochondrial membrane potential (ΔΨm) in cardiomyocytes. FoxO1 inactivation up‐regulated the expression of Kir6.1 and increased the ΔΨm in cardiomyocytes. Chromatin immunoprecipitation assay demonstrated that the Kir6.1 promoter region contains a functional FoxO1‐binding site. In conclusion, Kir6.1 improves cardiac dysfunction in DCM, probably through the AKT‐FoxO1 signalling pathway.

resulting in persistent FoxO1 nuclear localization and activation. 3,4 Our recent study showed that persistently high insulin levels result in a significant decrease in the expression of phosphorylated AKT (p-AKT) and FoxO1 (p-FoxO1), mitochondrial membrane potential (ΔΨm) and cardiac function in db/db mice, which indicates the links between altered insulin signalling and mitochondria in DCM. 5 ATP-sensitive potassium channel (K ATP ) plays an important protective role in the heart through various signalling pathways. K ATP activation protects cardiomyocytes during heart failure, decreases ischaemia/reperfusion injury and reduces the occurrence of arrhythmias. 6 K ATP is composed of two types of subunits, inwardly rectifying potassium channels and sulphonylurea receptors, and its subunit composition is tissue specific. [7][8][9] There is a K ATP channel in the inner membrane of mitochondria (mitoK ATP ), [10][11][12] which the inwardly rectifying potassium channel 6.1 (Kir6.1) is a part of mitoK ATP channels in cardiomyocytes. 6,8 A previous study has shown that the expression of Kir6.1 in heart mitochondria is significantly reduced in the mouse model of type 1 diabetes. 13 14 The mice were randomly divided into two groups: the DCM group was fed an HFD for 4 weeks, injected with STZ (100 μg/g of bodyweight) and then fed with HFD for another 12 weeks; the control group was fed a regular diet (D12450J, Research Diets, New Brunswick, NJ, USA) and injected with the same volume of vehicle (0.1 mol/L sodium citrate, Sigma-Aldrich). Mice were housed five per cage, with free access to food and water. The HFD (5.21 kcal/g) consists of 60% calories from fat, 20% carbohydrate and 20% protein and the regular diet (3.82 kcal/g) contains 10% fat, 70% carbohydrate and 20% protein. The data regarding food intake and caloric intake for each experimental group are listed in Table S1. Plasma glucose levels were measured at the beginning and 12 weeks after the STZ injection by a Contour glucose meter (Roche, Basel, Switzerland).

| Animal preparation and DCM model
Mice with a fasting plasma glucose of over 13.89 mmol/L were considered diabetic ( Figures S1 and S2). Mouse serum samples were analysed for Insulin using commercial enzyme-linked immunosorbent assay (ELISA) kits (Ray Biotech, Norcross, GA, USA) according to the manufacturer's instructions. The data of insulin sensitivity test for all animals are provided in Figure S3. The mice were killed with an anaesthetic overdose of pentobarbital (100 mg/kg of bodyweight, Sigma-Aldrich) injected intraperitoneally to obtain their samples.

| Primary cardiomyocyte isolation and cell culture
Primary cultures of neonatal rat ventricular cardiomyocytes (NRVMs) were prepared from hearts of 1-2-day-old Sprague Dawley rats, as previously described. 5 NRVMs were cultured in Dulbecco's modified Eagle medium (DMEM) containing 10% foetal bovine serum for 48 hours.

| Viral vector construction and transduction
A recombinant adeno-associated virus serotype 9 containing Kir6.1 (AAV-9) and a recombinant adenovirus encoding Kir6.1 (Ad-Kir6.1) were packaged by Shanghai HanBio Company (Shanghai, China). The AAV-9 capsid has previously been reported to show a modest preference for cardiac tissue in vivo. 15 The mice were randomized into two groups and injected with the null control virus (AAV-C, 2.70 × 10 11 GC/mL, 100 μL per mouse) or AAV-9 (3.97 × 10 11 GC/mL, 100 μL per mouse) via the tail vein before being fed standard rodent chow or an HFD. For the in vitro experiments, after 48 hours of cell culture, the medium was changed with fresh DMEM containing serum and NRVMs were transfected by adding adenoviruses expressing green fluorescent protein (Ad-C, Viral titre 1.58 × 10 10 PFU/mL) or GFP-fused Kir6

| Echocardiography
The mice were anaesthetized with 3% isoflurane and continuously monitored for heart rate, breathing and temperature. Transthoracic two-dimensional M-mode echocardiography was performed on anes-

| Brain natriuretic peptide measurement
Mouse serum samples and culture supernatants of cells were analysed for brain natriuretic peptide (BNP) using commercial enzymelinked immunosorbent assay (ELISA) kits (Ray Biotech, Norcross, GA, USA) according to the manufacturer's instructions.

| Histological analysis
Hearts were fixed in 4% paraformaldehyde solution, embedded in paraffin and sectioned (5 μm thickness). After dehydration, sections were stained with haematoxylin and eosin (H&E) and then viewed under a microscope (Olympus, Tokyo, Japan). For quantification, cell area measurements were performed on five similar sections, and 100 nucleated cells were randomly selected to measure the mean cell area.

| Apoptosis analysis
Hearts were fixed in 10% paraformaldehyde and embedded in paraffin. Paraffin-embedded sections were incubated at 60°C for 15 minutes, dewaxed and rehydrated. Heart tissue sections (5μm thick) were used for apoptosis detection with a TUNEL assay kit (Roche), as previously described. 4

| Transmission electron microscopy
Hearts were fixed in 2.5% glutaraldehyde overnight, followed by osmication and uranyl acetate staining, dehydration in alcohol and embedding in epoxy resin (Solarbio Life Science, Beijing, China).
Ultrathin sections were stained with uranyl acetate and lead citrate (Sigma-Aldrich). The sections were viewed and imaged under a transmission electron microscope (Hitachi, Tokyo, Japan).

| RNA isolation and quantitative real-time PCR analysis
RNA from heart tissue or NRVMs was extracted with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). cDNA synthesis was performed with a PrimeScript TM RT regent Kit (Takara, Kyoto, Japan). Quantitative real-time PCR (qRT-PCR) was performed in duplicate in a total reaction volume of 25 μL using SYBR-Green master mix (Takara) and conventional protocols. The primer sequences for qRT-PCR are listed in Table S3. Expression was normalized to that of the housekeeping gene, 36β4. Quantitative data was calculated using the comparative CT method. (1:300; rabbit polyclonal, ab39670, Abcam), GAPDH (1:2,000; rabbit monoclonal, ab181602, Abcam). The signal intensity was measured and analysed by Image J software, as previously described. 3 The expression of specific proteins was normalized to the protein expression of GAPDH.

| Oxygen consumption rate measurement
A Seahorse Bioscience XFe96 extracellular flux analyser was used to measure the oxygen consumption rate (OCR) in NRVMs using a previously reported protocol. 16 NRVMs were plated at 3000 cells per well in XF media supplemented with pyruvate (1 mmol/L), glutamine (2 mmol/L) and glucose (10 mmol/L; Sigma-Aldrich). Four independent OCR measurements were taken for each condition: baseline, and following the addition of oligomycin (1 μmol/L), FCCP (2 μmol/L), and antimycin A (0.5 μmol/L) plus rotenone (0.5 μmol/L; Agilent, Santa Clara, CA, USA). The protein concentration of NRVMs was determined for each well using a standard Bradford assay. Data were analysed by the Wave software and Report generator. fed with HFD for another 12 weeks; the control group was fed a regular diet and injected with the same volume of vehicle (0.1 mol/L sodium citrate). Littermate Kir6.1 wt/lox /MerCreMer (K-C) mice were used as controls.

| Chromatin immunoprecipitation assay
Chromatin immunoprecipitation assay was performed as de-

| Statistical analysis
Statistical analysis was performed with SPSS 17.0 software. Data are expressed as the mean ± SEM (standard errors). Comparisons of parameters between two groups were performed with unpaired Student's t-test. Comparisons of parameters among groups were determined by one-way or two-way ANOVA, followed by Tukey's post hoc test. P < .05 was considered statistically significant.

| Cardiac and mitochondrial function are decreased in DCM mice and in insulin-resistant NRVMs
The cardiac function of DCM and control mice was analysed by echocardiography. Compared with the control mice, the DCM mice showed a significant reduction in CO, LVEF, LVFS and HW/BW, and an increase in LVPW; d, thereby exhibiting DCM (Table 1 and Figure 1A). ELISA demonstrated that the BNP protein expression was significantly increased in the heart of diabetic mice ( Figure 1B).
The myocardial structure was examined by H&E staining. Diabetic hearts displayed structural abnormalities, including abnormal cellular structures, the existence of foci with necrotic myocytes and increased cardiomyocyte areas ( Figure 1C). The TUNEL assay was performed to examine apoptosis of cardiomyocytes. The proportion of apoptotic cells was remarkably increased in diabetic hearts compared with the controls ( Figure 1D)

| The expression of Kir6.1 is decreased in HFD&STZ-induced type 2 diabetic mice and in chronic insulin-resistant NRVMs
qRT-PCR analysis indicated that the Kir6.1 mRNA expression was significantly decreased in DCM hearts ( Figure 2A). Western blot analysis confirmed that Kir6.1 protein expression was reduced by 40% in DCM mice compared with the control mice ( Figure 2B).
In accordance with the in vivo results, the mRNA and protein levels of Kir6.1 in NRVMs were obviously decreased after chronic insulin stimulation ( Figure 2C and D). These results imply that Kir6.1 expression decreases with the decrease in cardiac function, suggesting it may play a role in DCM.

| Cardiac-specific Kir6.1 knockout aggravates cardiac dysfunction in diabetic mice
We used cardiac-specific Kir6.1-knockout mice to further study the role of Kir6.1 in DCM. qRT-PCR and Western blotting confirmed that the mRNA and protein levels of Kir6.1 in the heart of KO mice were significantly decreased compared with those in the control mice ( Figure 4A and B).
Kir6.1 deficiency increased the deterioration in cardiac function induced by HFD&STZ, as manifested by the reduction in CO, HW/ BW, IVST, LVEF and LVFS (Table 1 and Figure 4C). Kir6.1 deficiency in the heart resulted in higher BNP protein level in DCM mice compared with the control mice ( Figure 4D). Cardiac-specific Kir6.1 knockout also aggravated cardiac pathological changes in DCM mice, as demonstrated by the quantitative data of cardiomyocyte area determined by H&E staining ( Figure 4E). Furthermore, the apoptosis rate in cardiac-specific Kir6.1-knockout DCM mice was significantly increased compared with that in DCM mice ( Figure 4F).

| Effect of Kir6.1 on the AKT-FoxO1 signalling pathway in DCM
In the MK-2206&INS group, the ΔΨm was lower than that in the control group. However, in the INS group it was higher than that in the MK-2206&INS group ( Figure 6B and C).  Figure 6E).

| D ISCUSS I ON
In this study, we investigated the role and mechanism of Kir6. and four regulatory sulphonylurea receptor (SUR1 or SUR2) subunits. 27 However, the composition of mitoK ATP channels is still un-  In this study, we found that the cardiac function in DCM mice was decreased, including systolic dysfunction, increase in BNP, cardiomyocyte hypertrophy and apoptosis, and abnormal changes in mitochondrial structure in vivo. Additionally, we found increased BNP levels and reduction in the OCR in vitro. DCM and its associated mitochondrial dysfunction have been observed in ob/ob, db/db and HFD-fed mice. 34,35 Furthermore, cardiac tissue from Akita mice displayed swollen mitochondria, lacking a welldefined cristae structure along with decreased states 3 and 4 respiration and ATP synthesis. 36 Our data agree with many previous studies on cardiac dysfunction in rodent models of DCM. [37][38][39][40][41] However, in the current study, Kir6. Cardiac insulin signalling mediates cellular homeostasis by controlling substrate use, protein synthesis, autophagy and cell survival. 42 Physiologically, binding of insulin to IR activates IRS1 and IRS2 and the downstream PI3K-AKT pathways. AKT is required for cardiac growth, metabolism and survival, and its targets include p70S6K (protein synthesis), Glut4 (glucose transport) and FoxO1 (gene expression). 43 Briefly, insulin exerts its function through AKT activation, which in turn phosphorylates FoxO1. In cardiomyocytes, FoxO1 is involved in the control of many important properties such as cell growth, metabolic adaptation, cell apoptosis, autophagy and resistance to oxidative stress. 44,45 Impaired glucose uptake in the diabetic heart is often linked with reduced expression or activity of the downstream intermediates in the insulin signalling pathway. In this study, the levels of p-AKT and p-FoxO1 were markedly downregulated in DCM. Decreased cardiac basal and insulin-stimulated phosphorylation of AKT and FoxO1 is evident in diabetic mouse models. 46 In our previous studies, prolonged HFD feeding of mouse models impaired AKT activation and FoxO1 phosphorylation, which resulted in persistent FoxO1 nuclear localization and activation, 3,4 consequently leading to cardiac dysfunction. Furthermore, our recent study showed a reduction in the expression of p-AKT and p-FoxO1 and in cardiac function in db/db mice. 5 K ATP plays a key protective role in the heart through various signalling pathways.
Specifically, genetic manipulation of cardiomyocyte insulin signalling intermediates has demonstrated that partial cardiac function rescue was achieved by up-regulation of the insulin signalling pathway in diabetic hearts. 47 Similarly, a previous study has reported that the cardioprotective effect of K ATP occurs at least partially by regulating the AKT-FoxO1 signalling pathway, which in turn influences the expression of PGC-1α and its downstream target genes. 48 Our recent study also showed that opening of mitoK ATP increased the phosphorylation of AKT and FoxO1, but the effects of this opening were blocked by the specific AKT inhibitor, MK-2206. 5  Heart failure is the main cause of death in patients with type 2 diabetes, but the molecular mechanism of the link between diabetes mellitus and heart failure is not clear. Insulin resistance is a sign of type 2 diabetes. IRS1 and IRS2 are the main insulin signalling components that regulate cell metabolism and survival. Our previous studies have shown that IRS1 and IRS2 play important roles in controlling cardiac function, metabolism and homeostasis. And inhibition of cardiac IRS1 and IRS2 may be a fundamental mechanism for inducing heart failure. 4 A previous study has demonstrated that failing heart with coronary patency shows insulin resistance, glycogen deposition and asymmetrical myocardial hibernation due to microcirculatory dysfunction. 54 And diabetic cardiomyopathy is characterized by insulin resistance, chronic myocardial ischaemia and features of myocardial stunning/hibernation. Furthermore, another study showed that higher basal glycogen deposition was detected in Kir6.2 knockout heart. 55 Our previous study suggests that Foxo1 plays an important role in promoting diabetic cardiomyopathy and controlling β-MHC expression in the development of cardiac dysfunction. 3 The interaction between the AKT-FoxO1 signalling pathway and Kir6.1 may therefore play a potential myocardial metabolic role in the onset of heart failure in light of the abovementioned study.
In this study, the isoflurane MAC (%) is too high and may seriously affect the systolic function without cardiac injury. NRVMs were treated with insulin to induce the cardiomyocyte model of insulin resistance, which is not so representative as cardiomyocytes isolated from the mice model of DCM, because of the difference phenotype of neonatal cardiomyocytes from adult ones. We also did not investigate the detailed molecular mechanism of Kir6.1's action on the AKT-FoxO1 signalling pathway, which is another limitation of this study. Our future study will focus on the molecular mechanisms of Kir6.1's interaction with the AKT-FoxO1 signalling pathway in DCM.
In conclusion, our results provided in vivo and in vitro evidence that Kir6.1 improves cardiac dysfunction in DCM, probably through the AKT-FoxO1 signalling pathway. Moreover, the crosstalk between Kir6.1 and the AKT-FoxO1 signalling pathway may provide new strategies for reversing the defective signalling in DCM.

ACK N OWLED G M ENTS
This work was supported by the National Natural Science Foundation of China [grant numbers 81570349, 81200157].

CO N FLI C T O F I NTE R E S T
The authors declare that there are no conflict of interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.